A theoretical framework for weight-value set construction in land suitability assessment

When using the rating and weighting (RAW) method in land suitability assess ment, planners often substitute the theoretical weight-value range [0, 1] w ith a weight-value set, an ordered set of discrete values reduced from the continuous weight-value range. In constructing such a weight-value set, one needs to decide on how many weight values should be used (size of the valu e set) and where in the value range they should be placed (the choice of we ight values). At present, one must make these decisions and defend their cr edibility on the basis of 'professional judgments: as a theoretical framewo rk for such purposes is simply not available. Can a theoretical framework b e established so that one can make these decisions systematically? if so, i s it advantageous to use such a framework, instead of the existing methods, in constructing weight-value sets? In this paper these two issues are expl ored. I found that the theoretical weight-value range [0, 1] is actually sl iced into segments; that in producing land suitability scores, weight-value sets that are constructed from these segments function just as effectively as the weight-value range; and that a five-rile framework comprised of the se weight-value sets offers a higher level of efficiency and greater transp arency than existing methods.